Behavioural Brain Research 261 (2014) 1–7 Contents lists available at ScienceDirect Behavioural Brain Research j ournal homepage: www.elsevier.com/locate/bbr Research report I endeavor to make it: Effort increases valuation of subsequent monetary reward a,b,1 a,b,1 a,b a,b,c,d,∗ Qingguo Ma , Liang Meng , Lei Wang , Qiang Shen a School of Management, Zhejiang University, Hangzhou, China b Neuromanagement Lab, Zhejiang University, Hangzhou, China c Department of Economics, Faculty of Arts & Social Sciences, National University of Singapore, Singapore d National Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University, Beijing, China h i g h l i g h t s • Effort, in general, could enhance subjective valuation toward gain–loss outcome. • FRN and P300 represent modulated effect of varied efforts during outcome evaluation. • P300 also exhibits the valence effect of feedback at the late stage of evaluation. a r t i c l e i n f o a b s t r a c t Article history: Although it is commonly accepted that the amount of effort we put into accomplishing a task would exert Received 11 October 2013 an influence on subsequent reward processing and outcome evaluation, whether effort is incorporated Received in revised form as a cost or it would increase the valuation of concomitant reward is still under debate. In this study, 21 November 2013 EEGs were recorded while subjects performed calculation tasks that required different amount of effort, Accepted 25 November 2013 correct responses of which were followed by either no reward or fixed compensation. Results showed that Available online 3 December 2013 high effort induced larger differentiated FRN responses to the reward and non-reward discrepancy across two experimental conditions. Furthermore, P300 manifested valence effect during reward feedback, with Keywords: more positive amplitudes for reward than for non-reward only in the high effort condition. These results Outcome evaluation suggest that effort might increase subjective evaluation toward subsequent reward. Reward processing Event-related potential © 2013 Elsevier B.V. All rights reserved. Feedback-related negativity P300 1. Introduction addressed the effect of prior effort on the processing of the resulting reward. According to our common knowledge, there is a close relation- One prevalent view holds that effort has disutility, according ship between reward and effort, since we seldom get any reward to which effort itself carries a negative value or cost. This neg- without effort. Ecological theories assert that we would recognize ative connection between effort and reward is found in a series the role prior effort plays when we process the subsequent out- of theories, including social equity theory [3]. According to the come, due to the fact that a better comprehension would lead effort discounting principle, the net value of a reward would be to better decisions in the future [1,2]. Quite often, we may judge higher if it is comparatively easily obtained [4–6]. In effect, effort whether a potential payoff is worth the effort it requires or not. level stands as a reference point against the earned rewards, and There are two common propositions accounting for this effort more effort corresponds to a relatively higher reference point effect. However, up to now, only a few empirical studies have [7]. On the other hand, other previous literatures revealed that humans prefer conditioned rewards that are earned with greater effort [8]. It was discovered that actions taken beforehand could increase people’s valuation of the following reward [9–11]. Com- pared with windfalls, people have decreased willingness to spend ∗ Corresponding author at: Department of Economics, National University of money from earned gains [12], which supported the view of effort Singapore, Faculty of Arts & Social Sciences, AS2 Level 6, 1 Arts Link, Singapore valuation. 117570, Singapore. Tel.: +6583228139; fax: +86 57187995372. In recent years, there is increasing interest to probe into the neu- E-mail addresses: [email protected], [email protected] (Q. Shen). 1 These authors contributed equally to this work. ral mechanisms responsible for reward evaluation, which makes 0166-4328/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.bbr.2013.11.045 2 Q. Ma et al. / Behavioural Brain Research 261 (2014) 1–7 it possible for us to understand the role prior effort plays in the accompanying reward and subsequently enlarge the amplitude of processing of the subsequent reward in a direct manner. For exam- FRN discrepancy at the feedback stage [21]. ple, Hernandez et al. (2013) investigated how prior effort influences Another component is the P300, which is the most positive the valuation of different reward magnitudes using functional deflection in the 200–600 ms period after the presentation of feed- magnetic resonance imaging (fMRI) [13]. In their experiment, the back information, which typically exhibits its maximum magnitude subjects received monetary rewards of varied amount after they at parietal sites. Early research found that the P300 could encode successfully performed calculation tasks that were difficult, easy the motivational/affective significance of the stimuli [22]. In its or already solved. It was considered that as the difficulty of the extension to risky decision-making, it was found that the P300 is task increases, people would have to make more effort to solve the sensitive to the magnitude of reward, which is consistent with its problem. A forced donation stage was implemented subsequently, role to embody the subjective motivation toward stimuli in a gen- which caused a loss for the subject. Results showed that gain and eral manner [23–25]. With respect to the role of P300 in encoding loss magnitudes positively and parametrically modulate activation valence of the received outcome, early studies claimed that gain in NAcc and anterior insula respectively only after high effort, which loss difference has no impact on the P300 [24,25]. However, sev- suggested that increased effort leads to an accompanying increas- eral recent studies indicated that the P300 is also sensitive to the ing relevance of the outcomes. One possible limitation of this study valence of feedback, which responds more positively to positive is that reward and non-reward did not appear at the same stage feedback than to neutral and negative feedback [26–28]. of the experiment, which makes it difficult to directly compare In the present experiment, we applied ERPs to investigate the people’s perceptions to reward and non-reward with preceding integration of effort and outcome information in the human brain. effort. Our objective was to test for a neural correlate of effort valuation. To date, however, few studies have explored the temporal The calculation tasks were revised from the work of Hernan- dynamics of how effort shapes people’s valuation of subsequent dez et al. (2013) mentioned above [13]. We only recruited male reward. To investigate how such a cognitive process is imple- participants for this experiment, mainly due to their more con- mented in the brain, we adopted event-related potentials (ERPs) sistent performance in terms of accuracy rate. Each participant to explore temporal substrates of the evaluation of reward and was asked to solve a certain number of multiplication and addi- no-reward after effort. In ERPs studies concerning outcome evalua- tive operation tasks. Once they provided the correct answer to tion and reward processing, FRN and the P300 are the most widely a problem, they got 50% chance to win a fixed amount of mon- reported and examined ERP components. etary reward, and they would receive no reward under the rest Feedback-related negativity (FRN), which is the most negative 50% circumstances. This manipulation aimed to induce reward and deflection around the 250–350 ms period post-onset of feedback, non-reward at the same feedback stage, which differs from the shows maximal amplitude over medial frontal scalp locations. The previous fMRI experimental design. The subjects would not be com- amplitude of FRN is larger following negative feedback, which pensated for their effort if they gave a wrong answer or failed appears when facing an incorrect response, game failure, or mon- to arrive at a solution to the problem within the time limit. The etary loss [14–17]. electroencephalography signals were recorded from the subjects According to the reinforcement-learning theory of FRN [16,17], throughout the experiment. Such a paradigm allowed us to explore when people make a risky decision, the negative prediction errors how the previous effort affects subjects’ neural responses toward induced by the unfavorable outcome would facilitate the release of the subsequent reward and non-reward. dopamine in midbrain, which would reduce its inhibition function Considering that differentiated FRN (d-FRN) toward the to dopaminergic projection areas ACC. This would subsequently reward/non-reward divergence of the outcome reflects both the result in an increased deflection of FRN at the frontal area of scalp prediction error and the motivational/affective evaluation, we posit and vice versa. Such a mechanism could explain why the unex- that the reward/non-reward FRN discrepancy in observing the pected losses would elicit a relatively larger FRN as compared to resulting outcome of high effort would be more pronounced. Given gains, which has been well established in the past decade [18]. that P300 is a reflection of the motivational salience, we postulate Beyond that, in a recent study, Bellebaum et al. (2010) investigated that the P300 would loom larger in the high effort condition than the role of FRN through a probabilistic risky decision-making task in the low effort condition. and found that FRN is also sensitive to the violation of reward mag- nitude expectation, which indicated that, besides valence, the FRN 2. Methods could also represent the salience of the prior stimuli [19].